To date an effective treatment approach which results in short and long term positive outcome for children with Attention Deficit Hyperactive Disorder (ADHD) has not been established. The proposed study addresses major therapeutic issues in the management of children with ADHD. These include 1) determination of the relative short-term efficacy of different stimulant medications and of each subject's 'best' medication regimen; 2) comparisons of conventional stimulant treatment alone and in combination with full multimodal or single-modality psychosocial treatment; 3) how particular treatments affect the feasibility of discontinuing medication; 4) an assessment of the characteristics of medication (and other treatment) refusers, and of the efficacy of psychosocial treatment in this group; and 5) predictors of treatment outcome, including child, parent, family and demographic characteristics, and comorbid diagnoses. The study will treat a total of 720 ADHD/ADD children aged 7 to 9 at 4 different sites. Children with a variety of comorbid conditions such as oppositional disorder (ODD), conduct disorder (CD), specific academic skills disorders, anxiety or affective disorders, will all be included. The design includes an initial 16 week random crossover double-blind trial which will involve 4 weeks each of placebo, methylphenidate, dextroamphetamine and pemoline. In this way the optimal drug and dose for each individual child will be determined. Children on this optimal medication will then be assigned randomly to 5 different treatment groups: (1) A conventional stimulant treatment group; (2) A multimodal treatment group which will receive social and academic skills training and remedial tutoring for the child, a school based behavioral treatment program and parent training and counselling; (3) A school domain intervention group which will receive the school based behavioral program; (4) A child domain intervention group that will receive social and academic skills training and remedial tutoring; 5) A parent domain intervention group where the parents will receive parent training and counselling. Families who refuse medication will be randomly assigned to either a multimodal or an attentional control group. Intensive treatment will occur weekly for 19 months. This will be followed by a placebo challenge to determine if stimulant treatment can be discontinued, and how this is affected by the particular treatment received. We will thus be able to evaluate the efficacy of optimal medication plus either full multimodal or single domain (school, child, or parent) psychosocial treatments or conventional stimulant treatment. The effect of comorbidity and other psychosocial factors on outcome of various treatment approaches will be explored.